基于多源数据融合的高铁枢纽多模式换乘客流分担率估计

Estimation Passenger Transfer Demand Multimodal Split in a High-Speed Railway Hub Based on Multi-Source Data Fusion

摘要:

为实现高铁客运枢纽换乘客流分担率的精准辨识，研究结合多模式公共交通大数据，提出了一类基于广义出行链的高铁客运枢纽换乘方式选择模型。通过对不同公共交通方式换乘阶段的关联和融合，提取以高铁客运枢纽为端点的个体广义出行链，分析高铁客运枢纽换乘客流的时空分布特征；综合考虑高铁客运枢纽乘客的个体经济社会属性、主观心理因素及其个体出行特征对换乘方式选择行为的影响，并结合个体广义出行链的特征，基于多指标多原因（MIMIC）模型和多元logit（MNL）模型提出高铁客运枢纽乘客换乘方式选择模型。以北京南站实际数据作为模型的输入，得到了换乘客流方式分担率的估计值。比对分析估计值和真实值可知，误差在可接受范围内。

Abstract:

In order to accurately identify travel mode split rate of transfer passenger flow in high-speed rail hubs， a travel mode choice model is proposed based on a generalized trip chain model based on multimodal public transportation big data. Through the correlation and fusion of different public transportation modes in the transfer phase， the individual generalized trip chain with the high-speed rail hub as the origin is extracted. Then， the temporal-spatial distribution characteristics of the transfer flow in high-speed rail hubs are analyzed. Comprehensively considering the influence of individual economic and social attributes， and subjective psychological factors in combination with individual travel characteristics on transfer mode choice behavior， a transfer mode choice model for passengers in high-speed rail hubs is proposed based on multiple indicators and multiple causes （MIMIC） and multi-nominal logit（MNL） model. Taking the actual data of Beijing South Railway Station as the input， the estimated travel mode split rate of the commuting passenger flow is obtained. A comparison and analysis of ground true data indicates that the estimation error is within an acceptable range.